[ad_1]
Skill-based matchmaking was built to make competitive games feel legitimate. A new study says that balance can carry a hidden cost, because equal-skill contests can create the kind of losing streaks that push people out of the queue.
The research, published in Management Science, argues that game matchmaking works better when it looks beyond raw ability and accounts for how people react to recent wins, losses, and competitive patterns. In an analysis of 5.4 million Lichess matches, optimized matchmaking lifted engagement by 4% to 6% compared with conventional skill-based approaches.
The old skill-based matchmaking or SBMM rulebook now looks too blunt for platforms fighting to keep people around.
Why do fair matches backfire
SBMM pairs opponents near the same ability level to create balanced contests. On paper, that sounds like the cleanest answer. Nobody wants to get crushed by someone far better, and nobody learns much from steamrolling a beginner.
The problem is sequence. A single loss is part of competition, but a run of losses can make a player feel trapped in a system that keeps serving up frustration.
That is the behavioral gap standard matchmaking misses. Each round shapes the next decision, whether someone queues again, takes a break, or closes the game for the night.
How much engagement is at stake
The Lichess data gives the finding real weight. Across 5.4 million matches, the optimized system increased engagement by 4% to 6% over regular skill-based matchmaking. In theoretical scenarios, the gains reached as high as 50%.
For players, the change would be subtle. A better system would treat recent outcomes as part of the signal, then build matchups around the longer session instead of leaning only on ratings.
For game makers, those small percentages can add up fast. According to the Global Games Market Report, the global gaming market is projected to generate nearly $188 billion annually, so even modest retention gains can turn into real platform value.
When does retention go too far
Smarter matchmaking doesn’t give developers a free pass to manipulate the queue. It raises a harder trust problem, because competitive integrity and retention goals can pull in different directions when systems become harder to read.
That tension gets sharper around pay-to-win design. The research found that paid advantages can improve engagement under specific conditions by changing the skill mix, but it does not present that as a universal win. More engagement isn’t always a better experience.
The danger for studios is invisible matchmaking that players stop trusting. If developers move beyond pure SBMM, they’ll need to prove the queue still respects competition. The next version of matchmaking has to keep people playing without making the game feel managed against them.
[ad_2]
Source link
